The fate of the ileal pouch in patients developing Crohn's disease

Recent studies have suggested that a subset of patients with Crohn's colitis may have a favorable outcome after ileal pouch-anal anastomosis and have advocated elective ileal pouch-anal anastomosis in selected patients with Crohn's disease. We have not offered ileal pouch-anal anastomosis to patients with known Crohn's disease, but because of the overlap in clinical presentation of ulcerative colitis and indeterminate colitis, some patients receiving an ileal pouch-anal anastomosis are subsequently found to have Crohn's disease. We review our experience with these patients to identify potential preoperative predictors of ultimate pouch failure.

Patients with a final diagnosis of Crohn's disease were identified from an ileal pouch-anal anastomosis registry. These patients are followed prospectively. Preoperative and postoperative clinical and pathologic characteristics were evaluated as predictors of outcome. Median (range) values are listed.

Thirty-two (18 females) patients (4.1 percent) with a final diagnosis of Crohn's disease were identified from a registry of 790 ileal pouch-anal anastomosis patients (1980-2002). Patients underwent ileal pouch-anal anastomosis in two stages (11 patients) or three stages (21 patients). The preoperative diagnosis was ulcerative colitis in 24 patients and indeterminate colitis in 8 patients. Median follow-up was 153 (range, 13-231) months. The median time from ileal pouch-anal anastomosis to diagnosis of Crohn's disease was 19 (range, 0-188) months. Complications occurred in 93 percent, including perineal abscess/fistula (63 percent), pouchitis (50 percent), and anal stricture (38 percent). Pouch failure (excision or current diversion) occurred in nine patients (29 percent) at a median of 66 (range, 6-187) months. Two of these 9 patients had preoperative anal disease (not significant). Comparing patients with failed pouches (n = 9) to patients with functioning pouches (n = 23), post-ileal pouch-anal anastomosis perineal abscess (67 vs. 26 percent, P = 0.05) and pouch fistula (89 vs. 30 percent, P = 0.01) were more commonly associated with pouch failure. Preoperative clinical, endoscopic, and pathologic features were not predictive of pouch failure or patient outcome. For those with a functional pouch, 50 percent have been or are currently on medication to treat active Crohn's disease. This group had six bowel movements in 24 (range, 3-10) hours, with leakage in 60 percent and pad usage in 45 percent.

Patients who undergo ileal pouch-anal anastomosis and are subsequently found to have Crohn's disease experience significant morbidity. Preoperative characteristics, including the presence of anal disease, were not predictive of subsequent pouch failure. We choose not to recommend the routine application of ileal pouch-anal anastomosis in any subset of patients with known Crohn's disease.